Opening Address by SMS Jasmin Lau at GovTech STACK Conference 2026 Day 1

Good morning, and welcome to STACK 2026.

GovTech Singapore turns ten this year. Ten years ago, Singpass was mostly a password that people kept forgetting. Today it’s how most of us initiate transactions with government, often without even thinking about it. That is what good infrastructure does. After a while, you stop noticing it.

A lot of people built this infrastructure. Some are in this room, and some have moved on. Some are coming on board to make it better. I want to thank all of them.

Technology has changed a lot in ten years. What we are here to do has not changed, to make things better for the people and businesses we serve.

So let me start with a simple question, which I will come back to at the end. What problem are we trying to solve, and whose life gets better when we solve it?

It sounds very obvious. But it is the question we skip most often.

Let me give you a few examples from GovTech’s own work.

The first is not glamorous. It is about plumbing.

For years, almost every agency that wanted to do serious work with data had to build its own pipelines, its own storage and its own analytics. The same problems get solved again and again across different agencies, in a slightly different way each time. That’s expensive. It also requires talents to be spread across all of our agencies. When it came to sharing data, it also means data from one agency rarely fits neatly with data from another.

The Standard Data Platform changes that. GovTech provides the common pieces to bring data in, process it, store it and serve it. The shared standards also mean datasets can be joined up for analysis, while shared infrastructure reduces duplication and allows teams to focus on applying data to real problems.

EnterpriseSG went from onboarding the Standard Data Platform to production in just four months. They are now using it for trade intelligence and forecasting, and for tools that let the officers ask questions of enterprise-wide data in plain English.

Nobody writes a news story about plumbing. Similarly, nobody writes a story about a data platform. But it means the EnterpriseSG team now spends its time on enterprise problems, not on rebuilding data pipelines. Building good and shared foundations allows us to work on harder things.

The second example is about trust.

AI can now do things that would have been unimaginable a few years ago. People are right to ask hard questions about it. Can I trust what AI tells me? What happens to my data?

These are all fair questions, and the public service must be able to answer them. When something goes wrong, someone has to be accountable. And that someone in government is us - in GovTech.

Chat.gov.sg (opens in new tab) is one way we have tried to answer these questions.

Chat.gov.sg (opens in new tab) is a chatbot that helps Singaporeans find information and services across government in one place. It draws on trusted government information and when it does not have a reliable answer, it says so instead of making one up. It also respects clear boundaries. For instance, it can explain the Central Provident Fund Board's (CPFB) policies, but it will not recommend you which financial product to buy.

Before its launch, the team tested it on 50 citizen-facing topics, in the different ways Singaporeans actually talk, including in Singlish. Then they set automated red-teaming agents on it to try to trick it into breaking its own rules. Whatever got through was used to tighten the guardrails before launch. Since then, real usage and monitoring have shaped further improvements.

What I like about this is that the team spends as much effort asking how it might fail as asking how it might help. This helped us to strengthen the safeguards around it.

Now, this brings me to my third example. Even when we know what impact we want to create and how to pursue it responsibly, there are many complex problems that cannot be solved by a single product or organisation. They require collaboration across industry and government.

This third example is scams. This is one where clearly government cannot do it alone.

Scammers are good at what they do. They move across channels very quickly, and they understand very well how people behave when they are worried or in a hurry.

Take WhatsApp. All of us use WhatsApp daily. Encryption keeps our conversations private, which is what we want. But it also means no one else can see a suspicious conversation you are having with someone and then step in to warn you. The scammers know this.

GovTech is testing an app called AntiScam Pango. It checks messages on WhatsApp, Telegram and SMS, flags the ones that look suspicious, and warns you in real time, before you take further actions that you may regret later on. The analysis happens on your phone itself, so your messages continue to be private.

To make AntiScam Pango smarter, GovTech is working with Google DeepMind on more advanced on-device AI models.

GovTech brings years of anti-scam experience in our local context. Google DeepMind brings frontier AI capabilities. The models are trained on real and synthetic Singapore scam messages, so they recognise the language and tricks used on people here, not just generic scams from around the world.

Pango is now in open beta on Android. As it matures, we may look at bringing it into ScamShield.

Neither GovTech nor Google DeepMind could have built this alone. I hope it encourages others, especially the companies running the messaging platforms people use every day, to think about what they can also contribute.

This is really why STACK exists. To get people with different capabilities to learn from one another and work together to solve problems.

The problems worth solving today do not just stay inside one organisation’s boundaries.

This year, we are launching the first STACK Expo, transforming STACK into a full-scale industry experience where GovTech and our industry partners can showcase their latest solutions.

The government can bring clearly defined real-world problems. Industry and practitioners bring the expertise, and the honesty about what has and has not worked. Then we test things together and keep what works.

So please use these two days to find people to work with, especially on the problems you have trouble cracking on your own.

GovTech’s first 10 years were about building foundations. Many people worked very hard on them, often without much recognition, and the next ten years will stand on what we have built.

The technology will keep changing. Some of what you see this week will look dated in a few years. That’s fine. What should not change is the habit of asking: what problem are we solving, and whose life will get better?

If we keep asking that honestly and keep working with people who can help us answer it, I think the next 10 years will be good ones.

Thank you and enjoy STACK 2026.